OpenAeroStruct is a lightweight Python tool that performs aerostructural optimization of lifting surfaces using OpenMDAO. It uses a vortex lattice method (VLM) for the aerodynamics analysis and a spatial beam model with 6-DOF per element for the structural analysis.
Documentation is available here.
To use OpenAeroStruct, you must first install OpenMDAO 1.7.3 by following the instructions here: https://github.com/openmdao/openmdao. If you are unfamiliar with OpenMDAO and wish to modify the internals of OpenAeroStruct, you should examine the OpenMDAO documentation at http://openmdao.readthedocs.io/en/1.7.3/. The tutorials provided with OpenMDAO, especially the Sellar Problem, are helpful to understand the basics of using OpenMDAO to solve an optimization problem. Note that OpenMDAO 1.7.3 is the most recent version that has been tested and confirmed working with OpenAeroStruct.
Next, clone this repository:
git clone https://github.com/mdolab/OpenAeroStruct.git
Lastly, from within the OpenAeroStruct folder, make the Fortran files:
make
Note that the code will run without compiling the Fortran library, but it will run significantly faster when using Fortran.
run_vlm.py is for aero-only analysis and optimization. It can use a single lifting surface or multiple separate lifting surfaces.
run_spatialbeam.py is for structural-only analysis and optimization. It can use a single structural component or multiple structural components, where each component represents a spar within a lifting surface.
run_aerostruct.py performs aerostructural analysis and optimization.
For each case, you can view the optimization results using plot_all.py. Examine its docstring for keyword information.
An example workflow would be:
python run_aerostruct.py 1
python plot_all.py aerostruct.db
The first command performs aerostructural optimization and the second visualizes the optimization history.
The keywords used for each file are explained in their respective docstrings at the top of the file.
If you wish to examine the code in more depth, see run_classes.py and the methods it calls. These methods interface directly with OpenMDAO.
This current version of this repository has grown past the previous Matlab implementation. If you are looking for a Matlab-capable version, please see https://github.com/samtx/OpenAeroStruct for the latest version.
- The increase in accuracy of results when using a cosine-spaced mesh is not as large as it should be.
- Aerostructural optimization sometimes fails to converge for certain geometries. The example provided in
run_aerostruct.pyshould converge. The structural and aerodynamic values must make sense together, e.g. the beam thickness and radius must be able to support the aerodynamic loads.